Local emissions of pharmaceuticals at production sites can cause severe environmental problems. Currently there are no laws, regulations, or rules to demand the pharmaceutical sector to report emissions of pharmaceuticals from manufacturing facilities. There are however several voluntarily based initiatives both from governments, environmental organizations as well as from the pharmaceutical sector itself to monitor and promote less environmental pollution.
One such initiative is the model for environmental risk assessment of pharmaceutical production that was developed by IVL in 2019 (report B2352, Pålsson et al). This report describes the further development and testing of the model for the use in public procurement decisions in Sweden. The study also included an investigation of other, similar national and international initiatives as well as a roadmap for further development of the model.
The risk assessment of antimicrobial substances was further developed during the project with an inclusion of PNEC-R calculations that specifically address the predicted no effect concentration of resistance as opposed to the ordinary PNEC-value which is based on ecotoxicological effects within the recipient.
The PNEC-R calculation should reflect concentrations withing the wastewater stream leaving the manufacturing site. For antimicrobial substances both an ecotoxicological based PNEC and a PNEC-R value should be determined and the lowest of the two should be used in the following risk calculation. Furthermore, the PEC-calculation for antimicrobial substances should be adjusted to not include the removal rate (%) for off-site wastewater treatment, since the mixture of different wastewater streams in off-site treatment plants rather increases than decreases the development of antimicrobial resistance.
The intended testing of the model with real world data provided by the companies, was unfortunately not feasible within the project. The main reasons given by the companies for not submitting data were competing priorities in other project and internal reorganisations, competing priorities in global organisations or external contract partners that were unwilling to provide the requested information.
Nevertheless, the theoretical evaluation and testing showed that it will be possible to use the proposed model and the risk phrases it generates within public procurement decisions. The model could be incorporated into the already existing evaluation system in Fass and thereby provide the procurement organization with the necessary risk phrases without revealing confidential data outside the third-party reviewer team. This is most likely an important addition to the existing suggested package of procurement criteria by the National Agency for Public Procurement as there are no possibilities at hand to give preference to bidders from a performance point of view. Making use of the different risk phrases in public procurement can be a more elaborated and distinct way to evaluate the input from bidders.
Sidbrytning
The mapping of the national and international initiatives showed that there are many relevant ongoing projects and initiatives covering models for risk assessment on pharmaceutical production and/or environmental aspects in public procurement of pharmaceuticals. Cooperation and alignment of the model developed by IVL with the other initiatives is needed. Further need for development of the model are the following:
In a complex value chain, the final risk phrase for the product is represented by the site with highest risk phrase, no matter how well the rest of the sites performed. This makes the model somewhat blunt. In future versions of the model, it is desirable that this is adjusted to better reflect the environmental performance of the whole value chain.
The PEC-local calculation for antimicrobial substances could be further updated to better reflect the concentrations in the wastewater stream leaving the on-site treatment facility, where the PNEC-R is preferably applied, rather than the concentration within the recipient.
Alignment of the model with the industrial initiatives taken by e.g. EFPIA.